Basics GPUs Practice Reviews

PCIe 4.0 and PCIe 3.0 – Difference between x8 and x16 with the fastest cards – Where does the bottleneck begin?

Let’s have a look at the two tested cards in an overview and summarize the tested cards graphically, before I start the counter-test with other cards. Here first the behavior of the two bus types among themselves, because one sees very nicely the advantage of the PCIe 4.0, if the connection was halved to @8. All measured deviations are well below one percent, so that they can probably be neglected in reality.

I used other relevant NVIDIA cards for a simple plausibility test, but did not use the complex conversion with the water cooling. Of course I heated up all cards to the maximum before the runs, but already the circumstance of changing temperatures and the up to 2 runs per card and resolution resulted in almost 40 more benchmarks, which simply could not be done more extensively due to time constraints. Of course everything could have been done much faster with synthetic benchmarks, but nobody plays them.

The result is quite interesting, because even a GeForce RTX 2060 Super isn’t limited significantly from 1080p. The GeForce RTX 2070 Super almost reaches the 1 percent mark at 1440p and even the GeForce RTX 2080 Super or RTX 2080 Ti are virtually symptom-free at 2160p. But here too, with the addition of AA, a slight increase can still be seen, albeit not as strong.

Summary and conclusion

High-performance PCIe 3.0 graphics cards are used for example is quite limited on a B550 motherboard when the bus width is actually halved to @8. However, this happens to varying degrees depending on the game, resolution and quality settings. Even with the Quadro RTX 6000, things never really get dramatic. Only in 720p is a real slump measurable, only users with strong cards tend to play in higher resolutions.

Even the Radeon RTX 5700XT was never really caught up in bandwidth limits, so that at least AMD’s thought process of cutting the chipset can be understood under their own marketing aspects. Why should AMD still focus on PCIe 3.0 when the current products don’t require it anymore and new ones are already on the way? even an RX Vega64 is much too slow to run into any significant limits here. Only the Radeon VII is somewhat punished with the missing PCIe 4.0 functionality, but that’s almost niche.

Danke für die Spende



Du fandest, der Beitrag war interessant und möchtest uns unterstützen? Klasse!

Hier erfährst Du, wie: Hier spenden.

Hier kannst Du per PayPal spenden.

About the author

Igor Wallossek

Editor-in-chief and name-giver of igor'sLAB as the content successor of Tom's Hardware Germany, whose license was returned in June 2019 in order to better meet the qualitative demands of web content and challenges of new media such as YouTube with its own channel.

Computer nerd since 1983, audio freak since 1979 and pretty much open to anything with a plug or battery for over 50 years.

Follow Igor:
YouTube Facebook Instagram Twitter

Werbung

Werbung